Aolani–NVIDIA Blackwell Ultra Expansion Targets Malaysia AI Cloud

Aolani, an NVIDIA Cloud Partner since 2023, said on September 22, 2026 that it will deploy twenty-two thousand Blackwell Ultra GPUs with NVIDIA across new Malaysia and Philippines facilities beginning in early 2027, aiming to push its footprint past forty-eight thousand GPUs and one hundred megawatts. While the companies did not split capacity between the two markets, Malaysia’s Johor corridor and Klang Valley already host competing AI-factory narratives—from Firmus’s OpenAI-anchored contracted megawatts to YTL’s Sedenak gigawatt campus plans—so another Blackwell Ultra pool matters for startups that need liquid GPU-as-a-service rather than hyperscale tenancy alone.

Filed under Infrastructure and dated September 26, 2026, this AI4Malaysia briefing treats the Aolani expansion as Malaysian AI-cloud competition news distinct from yesterday’s Troops twelve-million-dollar orchestration venture. Revenue-sharing and credit-support language suggests capacity will chase signed demand, a useful counterweight to empty-hall risk as MIDA-scale approvals keep climbing.

Why it matters: Malaysian builders need choice beyond a few campuses. Extra Blackwell Ultra inventory can deepen markets—but only if power, interconnection and residency SLAs stay credible.

What it means in practice

Aolani–NVIDIA Blackwell Ultra Expansion Targets Malaysia AI Cloud — contextual photo

Malaysian procurement and startup leads should compare Aolani quotes against Firmus and telco clouds; demand named Johor or Klang Valley site timelines; assign an owner for credit-support contract review; run time-boxed inference proofs; and prefer designs that keep humans on spend caps. Place the expansion beside OpenAI’s Firmus Malaysia compute and YTL’s Sedenak campus plan.

Caveats come first. Undisclosed splits can leave Malaysia with a thin share; early-2027 starts depend on power delivery; and mid-tier clouds can struggle on margins. AI4Malaysia therefore presents the Aolani–NVIDIA plan as directional infrastructure context until Malaysian site filings appear.

What to watch next: first Malaysian campus address; energisation dates; and whether local ISVs become anchor tenants. Readers can continue on the AI4Malaysia homepage, or browse the Newsroom for additional briefings.

Bottom line: treat this update as orientation, not instruction. Malaysian AI infrastructure is thickening with another GPU cloud entrant and remains early. Organizations that benefit most will verify megawatts, keep humans on spend authority, and refuse to confuse a regional announcement with finished clusters.